User:KoshVorlon/WierdWiki/Durova
Headline text[edit]
Durova Part 1
-
- The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section.
- I have placed an indefinite block on this account as a disruptive sockpuppet. Due to the nature of this investigation, our normal open discussion isn't really feasible. Please take to arbitration if you disagree with this decision. Thank you. DurovaCharge! 16:48, 18 November 2007 (UTC)
- :I am sorry but some degree of further explanation here will be necessary. I appreciate your desire to keep your investigation techniques confidential but a first review of the contribution history here reveals no disruptive edits of any nature and no warnings of any kind. A response on an urgent basis is requested. Newyorkbrad (talk) 16:57, 18 November 2007 (UTC)
- ::What what? Yeah, Durova, you're really going to have to explain this. I see no transgressions of any kind on the part of this user; indeed, with over 100 DYKs, he seems to be a pretty positive force around here. GlassCobra 17:00, 18 November 2007 (UTC)
- :I agree with NYB... someone who writes 100+ DYK articles simply can't be here purely to be disruptive. Explain to NYB in private, there's no security issue there, and I think nearly everyone would drop this if he says in public "the secret evidence is enough". But as it is... I'll take this to ArbCom if no one else will, and I don't even really know User:!! --W.marsh 17:01, 18 November 2007 (UTC)
- :::Grossly poor block. I happen to know who this "disruptive" editor is, and I'll tell you now he's done absolutely nothing wrong at all Any admin who wants to know can email me. Majorly (talk) 17:04, 18 November 2007 (UTC)
- : Uh, yeah I'm sorry, there's got to be more transparency than this when we're talking about an editor with that contribution history. Even if we're talking about a good hand account that is looking for adminship at some point. ELIMINATORJR 17:10, 18 November 2007 (UTC)
- I will be more than happy to explain my research to the arbitration committee. Please take this there if that is your opinion. WP:AGF, please: I don't do something this bold without very good reasons. DurovaCharge! 17:06, 18 November 2007 (UTC)
-
- *(fourth edit conflict) As Durova is not a checkuser, I find it difficult to think of what could be confidential about the evidence in this case. Should not !! be unblocked until others have checked the situation? Sam Blacketer (talk) 17:12, 18 November 2007 (UTC)
-
- ::If the reasoning behind the block involved personal information of editors then this is the correct thing to do. If you truly object then make your objections known to arbcom. They are not just going to let Durova block someone for the fun of it. There are certain issues that cannot be transparent here, this is a fact of the wiki. I have seen it happen a few times and when I investigated it was indeed correct not to discuss it publicly. Unblocking without talking to arbcom first would be about as irresponsible as Durova making this claim without good evidence(which I am sure he has). 1 != 2 17:13, 18 November 2007 (UTC)
-
- :Given arbcom's current level of activity that isn't really an option. If you won't answer to the community chose half a dozen admins and explain your reseach to them.Geni (talk) 17:13, 18 November 2007 (UTC)
-
-
- I can understand where Durova is coming from. I don't find the suggestion to send this to ArbCom unreasonable, given Durova's history. Mercury 17:15, 18 November 2007 (UTC)
- ::This has been a tough call, but in my opinion a necessary one. I am very confident my research will stand up to scrutiny. I am equally confident that anything I say here will be parsed rather closely by some disruptive banned sockpuppeteers. If I open the door a little bit it'll become a wedge issue as people ask for more information, and then some rather deep research techniques would be in jeopardy. As I've said this before, take me to arbitration if you want to challenge this. I think I've said that enough times clearly - I opened this thread for exactly that purpose. More than half a dozen administrators have already seen this research. DurovaCharge! 17:16, 18 November 2007 (UTC)
- :::Names?Geni (talk) 17:19, 18 November 2007 (UTC)
- :::Are you willing to send your evidence to other, independent admins who haven't been cherrypicked? If so, click here. If not, then yeah, I'm sorry, this is going to ArbCom. Not out of an assumption of bad faith, but because such an extreme action with no public evidence must be confirmed by experienced editors. You say that more than 6 other administrators have "seen the research" but none other than you have commented supporting the block here, so there's a bit of a disconnect. —bbatsell ¿? ✍ 17:30, 18 November 2007 (UTC)
-
- Durova has made a reasonable request. Her block should be respected pending arbcom review. Tom Harrison Talk 17:21, 18 November 2007 (UTC)
-
- *Assuming good faith applies equally to !! as to Durova. Blocks are there to prevent disruption and I see no evidence that !! was being disruptive, or would suddenly start to be disruptive if unblocked. Sam Blacketer (talk) 17:22, 18 November 2007 (UTC)
- :Surely you can see the liability and danger in revealing investigative techniques? Mercury 17:24, 18 November 2007 (UTC)
- ::Security through obscurity? We've tried it a few times. has resulted in things like the main page being deleted and various inappropriate images turning up on the main page
- *Durova, could you please explain why you can't elaborate to the community and this has to go through ArbCom? Are there privacy concerns? Or is this soley to keep your techniques secret? If it is the latter then this is extremely disappointing. Ryan Postlethwaite 17:27, 18 November 2007 (UTC)
- :Can I suggest that rather than demanding things we can't have (names) or speculating, we wait for ArbCom? Durova is in direct contact with several arbitrators and CheckUsers at the moment, and her past investigation skills are held in some regard. Durova is not one to block lightly. Guy (Help!) 17:29, 18 November 2007 (UTC)
- ::Why can we have the names of the admins who have reviewed the evidence?Geni 17:31, 18 November 2007 (UTC)
- Agree with Tom Harrison and Mercury. We clearly have a sock infestation. If disclosing details would hamper future sock ID techniques, it is better for ArcCom to review it confidentially. If someone is truly concerned, take it there. Complaining about it here is counterproductive. Crum375 (talk) 17:30, 18 November 2007 (UTC)
- :::If there are arbitrators/checkusers who have evaluated the evidence it would be very helpful if they would comment here; after all, it would be counterproductive to start an ArbCom request if ArbCom is already reviewing the case. --Akhilleus (talk) 17:33, 18 November 2007 (UTC)
-
- AGF'ing both sides, while I see no evidence of disruption (as a DYK frequenter), I'm not going to say anything against either side since I don't know the underlying information. I would like to see the evidence if Durova trusts me with it, but if not then I understand. Iff Arbcom can resolve this in a imely matter then I don't mind it going there. Wizardman 17:35, 18 November 2007 (UTC)
-
- In lieu of my filing a formal request for arbitration, I urge that one or more arbitrators review this situation immediately. Based on the information available to me there is insufficient evidence to support any block. Newyorkbrad (talk) 17:37, 18 November 2007 (UTC)
-
- :Here's my question, if he's a sock, who's the master? Kwsn (Ni!) 17:43, 18 November 2007 (UTC)
- ::Hmm, I know of !! and the thousands of excellent contributions he has made... No idea what this about, can we have some idea as to the nature of the transgression at least? My impression was he'd exercised his right to disappear and return... The Rambling Man (talk) 17:46, 18 November 2007 (UTC)
- * I also don't get it...I understand not giving away confidential personal information, but could Durova at least cite the disruptive behavior for which this particular account was blocked? If it's a good hand/bad hand thing, then block the bad hand indefinitely and warn/short block the good hand. But this user account seems to be a productive one. Some of the recent sockpuppet hunting and blocking seems a little witch-hunty to me - what's the point of blocking alternate accounts that are not causing disruption? Videmus Omnia Talk 17:51, 18 November 2007 (UTC)
-
- Meta comment: Thank you, all, for not jumping to undo the block. Durova has unblocked !!, and I expect she will explain shortly. - Jehochman Talk 18:04, 18 November 2007 (UTC)
- === Unblock with apologies ===
- When I make a mistake I like to be the first to step forward to correct myself. It's very surprising that a few facts didn't come to light sooner, given the amount of time my report circulated and the people who had access to it. This is, in fact, not a first account. But it's a legitimate situation. I request early closure and archiving of this thread to protect that person's privacy. DurovaCharge! 18:06, 18 November 2007 (UTC)
-
Headline text[edit]
Part II
Question "courtesy blanking"[edit]
If an administrator makes a serious mistake, as Durova did here, does this fall under the "courtesy blanking" guidelines? I thought courtesy blanking was to protect the identities or other personal information of editors, not to put an admin's mistake out of plain sight. Is this a common practice and, if so, doesn't it threaten our goal of promoting transparency and accountability for our actions? Cla68 (talk) 02:17, 19 November 2007 (UTC)
- Personally, I think hiding away discussion like this only gives trolls the recognition they crave by seemingly confirming their claims of conspiracy and secrecy, instead of simply being open about everything that has gone on and heading off silly conspiracy theories at the pass. Also, if the "courtesy" is being extended to !!, it would probably be most courteous to leave the discussion public (though closed/archived), so that his innocence in the matter is made clear. --krimpet⟲ 02:22, 19 November 2007 (UTC)
- (edit conflicted)My request was specifically to courtesy blank for that editor's privacy so that any doubts about my own actions could continue to be discussed on their own merits, without collateral damage.[1] I'm not sure that what happened really was all that serious a mistake. Blocks get overturned all the time. I overturned my own action in 75 minutes, opened the thread here myself to invite scrutiny, and extended prompt apologies. The action itself turned out to be mistaken, but was neither hasty nor superficial. Nobody bats .1000. And I wish to emphasize (in case any lingering doubt exists), the editor I blocked is very much a productive individual in good standing. DurovaCharge! 02:24, 19 November 2007 (UTC)
- You're saying that it was courtesy blanked out of concern for the wrongfully-blocked editor? Cla68 (talk) 02:31, 19 November 2007 (UTC)
- Yes, out of respect for his privacy. I requested it and other editors agreed. I specifically opened my own actions to scrutiny twice: first in starting this enormous thread and then in requesting that my own behavior remain under scrutiny after his had withstood it. From your own opening post to this subthread, it seems you agree that was the right thing to do. DurovaCharge! 02:41, 19 November 2007 (UTC)
- What privacy? You killed any expectations of privacy with sleuthing. Not to mention there were no private details revealed above. ⇒SWATJester Son of the Defender 04:53, 19 November 2007 (UTC)
- Echoing Swatjester's rationale, I don't agree that courtesy blanking was the right thing to do here. For Durova to say that it was out of respect for the wronged editor is laughably disingenuous. I'd like to hear Crum375's explanation for why he blanked the discussion. Cla68 (talk) 06:08, 19 November 2007 (UTC)
- The only person such a blanking helps is Durova herself. east.718 at 06:31, November 19, 2007
- Echoing Swatjester's rationale, I don't agree that courtesy blanking was the right thing to do here. For Durova to say that it was out of respect for the wronged editor is laughably disingenuous. I'd like to hear Crum375's explanation for why he blanked the discussion. Cla68 (talk) 06:08, 19 November 2007 (UTC)
- What privacy? You killed any expectations of privacy with sleuthing. Not to mention there were no private details revealed above. ⇒SWATJester Son of the Defender 04:53, 19 November 2007 (UTC)
- Yes, out of respect for his privacy. I requested it and other editors agreed. I specifically opened my own actions to scrutiny twice: first in starting this enormous thread and then in requesting that my own behavior remain under scrutiny after his had withstood it. From your own opening post to this subthread, it seems you agree that was the right thing to do. DurovaCharge! 02:41, 19 November 2007 (UTC)
- You're saying that it was courtesy blanked out of concern for the wrongfully-blocked editor? Cla68 (talk) 02:31, 19 November 2007 (UTC)
- (edit conflicted)My request was specifically to courtesy blank for that editor's privacy so that any doubts about my own actions could continue to be discussed on their own merits, without collateral damage.[1] I'm not sure that what happened really was all that serious a mistake. Blocks get overturned all the time. I overturned my own action in 75 minutes, opened the thread here myself to invite scrutiny, and extended prompt apologies. The action itself turned out to be mistaken, but was neither hasty nor superficial. Nobody bats .1000. And I wish to emphasize (in case any lingering doubt exists), the editor I blocked is very much a productive individual in good standing. DurovaCharge! 02:24, 19 November 2007 (UTC)
Hi, does User:!! want it blanked? It may be a naive question, but it would seem this should be his decision. Has anyone asked him? • Lawrence Cohen 06:13, 19 November 2007 (UTC)
- He wasn't available at the time when I requested the blanking, but it seemed like the right thing to do. From the tone of his comments afterward I see no reason to question that. Rather than reopen that, please refactor the portions of this thread that pertain to him and concentrate on my actions if there's any more to be said. He's no longer under scrutiny. Whether I ought to be is not for me to decide. DurovaCharge! 06:31, 19 November 2007 (UTC)
- Sorry, wasn't trying to cast any aspertions either way. I was just thinking the whole question of blanking could be wrapped up immediately if he said a simply yay/nay to that. From reading all this the only thing I was curious about was the unanswered question (I saw it asked several times, by various folks, including !! on his talk page) of who saw the evidence. • Lawrence Cohen 06:33, 19 November 2007 (UTC)
Followup[edit]
It seems to me that this was a colossal failure to assume good faith, and that Durova has forgotten that this is an encyclopedia, not a practice ground for sleuthing. This is not the first time this has happened: "Sleuthing" evidence that Durova was not prepared to bring to the public drove Alkivar from the project. Durova, I would like to see some statement from you that you are going to stop accusing editors and blocking them without presenting any public evidence, and I mean "public" as in fully transparent, not only to those you deem worthy. ⇒SWATJester Son of the Defender 19:22, 18 November 2007 (UTC)
- I have to go along with this. Even if this user is Willy on Wheels, Karmafist, myself, or the Devil incarnate, there would be no reason to block until the account actually did something disruptive (such as create abusive sockpuppets, run at RFA under false pretences, or something). This appears to have been a botched attempt at killing a goose that's laying really splendid golden eggs (ouch, that's not hugely polite, my apologies to !!, but you get my point). At the very least we need reassurances that This Will Not Happen Again. Moreschi If you've written a quality article... 19:42, 18 November 2007 (UTC)
- ""Sleuthing" evidence that Durova was not prepared to bring to the public drove Alkivar from the project." Really? I thought he left because a RfA found against him and desysopped him. Darkson (Yabba Dabba Doo!) 19:48, 18 November 2007 (UTC)
- He left before the decision was rendered, because of the evidence. ⇒SWATJester Son of the Defender 19:49, 18 November 2007 (UTC)
- If the evidence was against him, and the desysopped was likely (as it seemed to be), why does it matter if he left before the RfC was closed? The evidence against him seemed pretty thorough, with or without Durova's. Darkson (Yabba Dabba Doo!) 19:56, 18 November 2007 (UTC)
- Agreed with Darkson here. Alkivar was going to be desysopped with or without the *puppetry allegations, and it should have happened months (if not years) before it did. —bbatsell ¿? ✍ 20:00, 18 November 2007 (UTC)
- He was going to be desysopped. He wasn't going to be driven from the project, as is where we are now. ⇒SWATJester Son of the Defender 20:06, 18 November 2007 (UTC)
- Sorry, but that's naive. Alkivar was taking his cues from a banned use,r probably the most prolific and disruptive sockpuppeteer currently active. Durova's evidence was not made public but was presented by email to arbitrators, and the arbitrators reviewed it independently. He left because he knew what was coming. I regret that, we probably all do, including Durova, but it's useless to blame Durova for the fact that Alkivar was dancing to Barber's tune, any more than it's her fault that Burntsauce was doing so. Guy (Help!) 22:26, 18 November 2007 (UTC)
- Agreed with Darkson here. Alkivar was going to be desysopped with or without the *puppetry allegations, and it should have happened months (if not years) before it did. —bbatsell ¿? ✍ 20:00, 18 November 2007 (UTC)
- If the evidence was against him, and the desysopped was likely (as it seemed to be), why does it matter if he left before the RfC was closed? The evidence against him seemed pretty thorough, with or without Durova's. Darkson (Yabba Dabba Doo!) 19:56, 18 November 2007 (UTC)
- He left before the decision was rendered, because of the evidence. ⇒SWATJester Son of the Defender 19:49, 18 November 2007 (UTC)
- ""Sleuthing" evidence that Durova was not prepared to bring to the public drove Alkivar from the project." Really? I thought he left because a RfA found against him and desysopped him. Darkson (Yabba Dabba Doo!) 19:48, 18 November 2007 (UTC)
- I'm with SWATJester 100% here. I don't think you can go and block a respected editor indef, announce it on AN/I then refuse to comment on any evidence, realise you've messed up and offer a two line apology and expect people to be satisfied here. Revealing your investigation techniques doesn't strike me as being a very good reason to direct everyone to ArbCom rather than allowing transparency in your block. Ryan Postlethwaite 19:49, 18 November 2007 (UTC)
- Yeah... You can't tell people that you MUST be taken to Arbcom or you will refuse to do anything. If there's some reason that only Arbcom can be given the evidence, then go to them BEFORE you block, don't block and make other people go through Arbcom to challenge it. -Amarkov moo! 19:54, 18 November 2007 (UTC)
- She sent her "report" to ArbCom before blocking !!. She says she got approval from ArbCom members to block the user as a violation of WP:SOCK. Nishkid64 (talk) 19:59, 18 November 2007 (UTC)
- Unfortunately, it was precisely these assumptions that allowed Burntsauce to cause serious damage to the project for half a year. Dannycali was also indeffed as a long term sockpuppet. Such things do exist, and I pledge to work on better checks and balances to reduce false positives such as this one in the future. My report had been in circulation for two weeks among some very senior people and I acted on the reasonable belief that any false positives would have come to light before this. If there's anything more to be said on the matter, please do this editor the courtesy of archiving this discussion out of respect for his privacy and focusing any criticisms on my actions alone. He shouldn't need to be a part of this. DurovaCharge! 20:01, 18 November 2007 (UTC)
- It is precisely relevant to this discussion. ⇒SWATJester Son of the Defender 20:04, 18 November 2007 (UTC)
- However, !! apparently did little but work hard on articles. I've seen no evidence that a single edit he made was considered disruptive... whereas with people like Burntsauce, who made some productive edits, you never had to look too far to find disruptive edits too. Where is evidence of any of this for User:!!? I apologize if I've missed it. --W.marsh 20:07, 18 November 2007 (UTC)
- We keep hearing about these "senior editors". But WHERE are they? Why don't any of these senior editors come forward and confirm this? -Amarkov moo! 20:05, 18 November 2007 (UTC)
I'll go ahead and say what I've heard plenty of people saying on IRC: It's disturbing that an editor so heavily involved in the SEO field, would use some "proprietary" investigative techniques on Wikipedia, ruin people's Wikipedia experience, and then refuse to provide any evidence to support the allegations out of a fear that those uber-valuable methods will become public. I don't find that acceptable in the slightest. I'd like to hear what Durova has to say about this, and what her plans are for future sleuthing. Forget about "pledging to reduce false positives". Why are you sleuthing in the first place? Why, as it appears, is it your mission to hunt down other editors using private evidence? ⇒SWATJester Son of the Defender 20:04, 18 November 2007 (UTC)
- To Swatjester: bizarre conversations like that are one reason why I oppose admin channel IRC on principle. DurovaCharge! 20:09, 18 November 2007 (UTC)
- That's not relevant and it won't wash. If we lose !! because of this it will take a good deal more than a few lines of rather weaselly apology for collective anger to be assuaged. I have great respect for you, Durova, but I think this shows we need a change of method. Somewhere wires got crossed - we need to make sure that doesn't happen again. Moreschi If you've written a quality article... 20:17, 18 November 2007 (UTC)
- While dissociating myself from the more extreme implications in SWATJester's report of what's being said on IRC, and acknowledging that some of the motivation for your actions have been spelled out in your candidate statement for the ArbCom elections, I'd like to see either a statement about what extra checks and balances you intend to employ if you continue your activity in this direction, or at the very least an acknowledgment that public distress in this case seems to indicate that you need to sharply reverse course in terms of sock-hunting. Relata refero (talk) 20:24, 18 November 2007 (UTC)
(edit conflicted)Something in Nishkid's statement requires correction: I did not send the report to ArbCom as a body. I did circulate it in ways that some arbcom members saw it. Nor do I say I got specific approval from ArbCom members to block: I circulated a report that roughly two dozen trusted people saw and no one objected. Please, Nish, clarify these things before speaking on my behalf. This puts me in an awkward position regarding sensitive information. If such posts continue I will only point out that some representations are not my words. DurovaCharge! 20:06, 18 November 2007 (UTC)
I agree with many of the critical points raised above, and in addition the attempt partially shift the blame to nameless senior people is pretty poor form. RxS (talk) 20:09, 18 November 2007 (UTC)
- It's pretty poor form to characterize things that way. At any rate, if there are serious concerns about my conduct and discretion I have no objection to having my actions scrutinized by people who have full access to the facts. Either ArbCom or the Foundation would be appropriate. DurovaCharge! 20:14, 18 November 2007 (UTC)
- Or, you could just explain who these nameless "senior editors" ARE, so that they can confirm what you're saying. -Amarkov moo! 20:17, 18 November 2007 (UTC)
- I think it's a pretty easy characterization: I circulated a report that roughly two dozen trusted people saw and no one objected. In addition you seem to be saying that outside of this 2 dozen people you won't allow your actions to be scrutinized, that's not the way things work here. RxS (talk) 20:25, 18 November 2007 (UTC)
- A more specific summary of the due diligence you conducted would reveal no private information at all but would allow the community a much better understanding of the procedure you undertook in making this block. I find it impossible to believe that there is a compelling reason not to provide this detail (i.e. on date I sent a summary of my investigation to names, who responded positively on date).Christopher Parham (talk) 20:19, 18 November 2007 (UTC)
- Isn't this case supposed to be closed as resolved, why are we still talking. This is a Secret account 20:21, 18 November 2007 (UTC)
- The issue is manifestly not resolved, in that a serious error was made and we have no understanding of why it was made or what will be done to prevent exactly the same mistake from being made again. Christopher Parham (talk) 20:23, 18 November 2007 (UTC)
- (ec) More to the point; that !! had a more than pristine record, with a history of DYK involvement and that this was instrumental in their unblocking. It just looked severely wrong and people were willing to express their shock and surprise here. Had it been some nameless account with a mediocre past, I'm sure the outcome would have been very different and we wouldn't be discussing it. They would have been gone - QED - with no recourse open to them. I'd really like to know a little more about what went wrong here so we can ensure it doesn't recur. Not everyone is afforded the level of 'justice', if you like, that this editor received - Alison ❤ 20:29, 18 November 2007 (UTC)
- A review of the Burntsauce and Dannycali histories would shed some light on that. It's not much of a secret that some banned editors do build up sockpuppets for long term use and attempt to mimic legitimate Wikipedians by doing some useful work. That's how, in spite of their disruption, both accounts survived for over half a year. They might still be editing if I hadn't examined them. What happened in this investigation was a rather odd set of coincidences lined up. DurovaCharge! 20:37, 18 November 2007 (UTC)
- That's perfectly understandable, but I hope you realise in the absence of any actual ability to scrutinize your methodology, the community is justifiably concerned. Simply put, the basis of any admin action must be satisfactory and timely review; otherwise, however much your intentions might be trusted by all of us, we will always have doubts about your unilateral actions. Relata refero (talk) 20:52, 18 November 2007 (UTC)
- A review of the Burntsauce and Dannycali histories would shed some light on that. It's not much of a secret that some banned editors do build up sockpuppets for long term use and attempt to mimic legitimate Wikipedians by doing some useful work. That's how, in spite of their disruption, both accounts survived for over half a year. They might still be editing if I hadn't examined them. What happened in this investigation was a rather odd set of coincidences lined up. DurovaCharge! 20:37, 18 November 2007 (UTC)
- (ec) More to the point; that !! had a more than pristine record, with a history of DYK involvement and that this was instrumental in their unblocking. It just looked severely wrong and people were willing to express their shock and surprise here. Had it been some nameless account with a mediocre past, I'm sure the outcome would have been very different and we wouldn't be discussing it. They would have been gone - QED - with no recourse open to them. I'd really like to know a little more about what went wrong here so we can ensure it doesn't recur. Not everyone is afforded the level of 'justice', if you like, that this editor received - Alison ❤ 20:29, 18 November 2007 (UTC)
- The issue is manifestly not resolved, in that a serious error was made and we have no understanding of why it was made or what will be done to prevent exactly the same mistake from being made again. Christopher Parham (talk) 20:23, 18 November 2007 (UTC)
- Isn't this case supposed to be closed as resolved, why are we still talking. This is a Secret account 20:21, 18 November 2007 (UTC)
How's this for a resolution? In the future I'll send such reports to the Committee formally and let them act. And if I happen to be on the Committee I'll let another member act. I don't want to create drama and I respect consensus. DurovaCharge! 20:24, 18 November 2007 (UTC)
- If you truly don't want to cause drama, why can't we hear one of these people you discussed the block with corroborate what you've said? That would pretty much kill the drama. -Amarkov moo! 20:28, 18 November 2007 (UTC)
- Well if I'd been a better dramatist I certainly would have lined up some people to me-too this thread. Hadn't anticipated the necessity. That's not my style. DurovaCharge! 20:31, 18 November 2007 (UTC)
- Seems you are doing OK on the drama front. More to the point. I'm sure that you blocked in good faith and that you did seek lots of feedback from sensible admins before blocking. The worry is that this might happen again and that's why so many people are fired up about this. Spartaz Humbug! 20:36, 18 November 2007 (UTC)
- Well if I'd been a better dramatist I certainly would have lined up some people to me-too this thread. Hadn't anticipated the necessity. That's not my style. DurovaCharge! 20:31, 18 November 2007 (UTC)
- Sounds like a good idea. To do otherwise, as you did, makes it look like unilateralism. I'm sure that's not the case, but it certainly looks that way. And yeah, the flipside is that ArbCom is so bunged up anyway that it could be some time before they act upon it - Alison ❤ 20:36, 18 November 2007 (UTC)
- It's a problem we need to get better at addressing as a site: some banned users refuse to go away, and not all of the sockpuppets they create are throwaway accounts. We're dealing with a situation here where a core of people share tactics and operate as a team. Probably most people who read the noticeboards regularly get that impression. One of the things I've been working on doing is to reverse engineer these people's playbook in a way that lets us identify them and act upon the problem. Such accounts are simpler to identify than to address because they do look legitimate to a superficial browse, and because they'd likely write a better playbook if they knew where their mistakes are. I do my very best to avoid false positives and I pledge absolutely to correct myself as quickly as possible when I make a mistake. In the future, situations where a substantial part of the evidence needs to be confidential are situations I'll route through ArbCom. I hope that satisfies reasonable concerns. DurovaCharge! 20:48, 18 November 2007 (UTC)
- (ec)Routing through ArbCom is a sensible first move; some scrutiny is better than none. I'd just say that I'd like to know that the actual "reverse-engineered playbook" has been scrutinized by those who are trusted by the community to ensure that drama of this sort is minimized. Relata refero (talk) 20:57, 18 November 2007 (UTC)
- It strikes me that this was a similar situation as what happened with Melsaran. Evidence was uncovered as part of an investigation and it was reported to ArbCom. However, given the potentially confidential nature of the evidence and the fact that there was no ongoing disruption, would it not have been best for an Arbitrator to make the block, specifically stating that they were doing so for the ArbCom, based on confidential evidence, as that is part of the reason ArbCom exists (to deal with such evidence)? Mr.Z-man 20:54, 18 November 2007 (UTC)
- You're right, and I realize that now. Will do. That was how we handled Runcorn/Poetlister, for instance. DurovaCharge! 20:58, 18 November 2007 (UTC)
(blanked post by sock of Amorrow)
- The key difference with Melsaran was that the evidence was supplied to ArbCom and they took the decision, with the result that the block stuck despite concerns over it. Sam Blacketer (talk) 20:59, 18 November 2007 (UTC)
With respect to: This account has been blocked indefinitely for violations of WP:SOCK. [2] the WP:SOCK page states:
The general rule is: one editor, one account. Do not use multiple accounts to create the illusion of greater support for an issue, to mislead others, to create controversy, or to circumvent a block. Do not ask your friends to create accounts to support you or anyone else. Multiple accounts are not for collusion, evasion, disruption, or misuse.
Would it be possible in the future to be more specific on what part of the policy is being violated? (Among creating the illusion of greater support for an issue, misleading others, creating controversy, or to circumventing a block.)
Of course I am not requesting the revealing of proprietary sleuthing techniques or any other information that the community at large need not be privy to. Uncle uncle uncle (talk) 21:23, 18 November 2007 (UTC)
- Durova, I wish you'd discussed it with me, though that was just bad luck. But is it necessary to change the subject above to the point of using terms like "banned users", " throwaway accounts, "share tactics and operate as a team", "these people's playbook", "look legitimate to a superficial browse", in a thread about the blameless User!! ? Don't you see how it makes a certain guilt by association stickily adhere to him? I hope you know by now how little that editor deserves such an aura. A good strong apology (I hve a low opinion of the apologies you did post) would be a lot better than irrelevancies about evidence needing to be confidential (qué? evidence of?) and how "we" need to get better at finding abusers. What does any of that have to do with User:!! ? Bishonen | talk 22:00, 18 November 2007 (UTC).
- When I first reversed the block I made an apology to this board and requested that the thread be closed out of respect for that editor's privacy. When the thread reopened I promptly requested courtesy blanking of that older portion so that any concerns could address my actions alone. I also delivered a prompt apology at that editor's user talk and reversed the block exactly one hour and fifteen minutes after implementation, which was as swiftly as I could verify the new information that came to light. After speaking to Bishonen by chat I have also archived my own user talk as a courtesy to this editor and, via a different intermediary, I have offered to communicate with this person by either e-mail or chat client, although I doubt that much more could be added to the corrective action and apologies I've already extended. Sockuppet investigations and blood tests sometimes yield false positives. Even checkuser results can yield false positives. About two months ago an administrator got blocked due to a false positive checkuser. That instance took longer than this to correct and the individual who performed the block did not reverse themselves with apologies. If this person informs me himself of other things I can do to set things right I'll certainly do all that I can within reason, but I have received no direct reply to these overtures. In the absence of any direct reply, I think the measures I've already taken are more than reasonable. DurovaCharge! 22:25, 18 November 2007 (UTC)
- Agree completely. I saw the evidence, it was definitely suspicious, but I can see where it went wrong. Unfortunate, but there does not seem to be any lasting damage, the editor was interrupted only briefly and a sincere apology has been given. What more are we supposed to do now? Guy (Help!) 22:29, 18 November 2007 (UTC)
- You're supposed to stop doing things like this. Nothing you in particular have done in terms of blocking recently has reduced drama at all. Relata refero (talk) 05:56, 19 November 2007 (UTC)
- Nothing. Durova has apologised and taken a reputation hit that means it is unlikely to happen again. Other than slight curiosity as to the identity of the half dozen admins I think we have reached case case closed move on and go find some copyvios or something.Geni 22:39, 18 November 2007 (UTC)
(blanked post by sock of Amorrow)
- Perhaps you missed part of the discussions. I had assembled a seven point report with 28 diffs and had circulated it for two weeks before acting. Obviously the methodology needs improvement, but the fact that it ultimately proved to be mistaken doesn't mean the approach was either hasty or superficial. I've pledged specific improvements so this doesn't happen again. DurovaCharge! 01:42, 19 November 2007 (UTC)
- The one thing that remains to be seen, in my view, is what Durova has done to correct an apparently flawed methodology for uncovering sockpuppets. When a blood test has the potential for false positives, doctors don't use it as the sole basis for initiating potentially devastating treatment in situations that are not time critical. My issue with Durova's response to date is that it casts this as the inconvenient byproduct of an effective process rather than as the result of a flawed process which requires correction. Additional oversight by the ArbCom is not likely to be a solution -- oversight by a handful of experienced editors was not sufficient to catch the errors in this instance, why would that be sufficient going forward? Christopher Parham (talk) 23:14, 18 November 2007 (UTC)
The question here isn't "was Durova wrong to have blocked !!", we already know the answer to that. The question here is why are we supporting a continuous failure to assume good faith, and what place does "sleuthing" have on this project? I thought we were here to write encyclopedias, not to dig up dirt on other editors, certainly not to character assassinate them with evidence that won't be made public by someone heavily involved in the SEO industry using their proprietary tactics. That's just wrong to me. We don't answer to them, we answer to the community, and the community is obviously NOT O.K. with private sleuthing being used without revealing the evidence and the methods involved. ⇒SWATJester Son of the Defender 23:08, 18 November 2007 (UTC)
- Responding to a couple of the recent posts, the corrective measures I'm taking have little to do with whatever effect today's events have on my reputation. Editors who contribute legitmately shouldn't be put on the spot. We all want to minimize that and I'll do my part, although not to the extreme extent one or two people suggest of giving up sockpuppet investigations. In a recent conversation about the 500+ JB196 socks I mentioned that I'd really like to be spending more time on other things, but the net good to the project of rooting out one Burntsauce probably exceeds the net good of creating one featured article: the damage being done by that individual was really quite extensive. Other than that, the points being raised at this stage of the discussion are repetitions of things that have already been addressed quite a few times to the satisfaction of most people. I'm sure the sockpuppeteers and their supporters would be delighted if fewer investigations were performed, but there's no need for such a radical and detrimental solution. DurovaCharge! 23:52, 18 November 2007 (UTC)
- I think Durova has the ability to be fair and reasonable. She blocked me previously for using two other accounts, which admittedly I did, but after I demonstrated that I could operate constructively with only my main account, seek mentorship through adopt-a-user, check in with her, alter various other editing habits, etc. she has allowed me to be unblocked and since being unblocked I have received some barnstars and much positive experiences with other editors.. Since being unblocked I have also unfortunately experienced some on-Wiki harrassment from a few accounts (two of which, thus far, checkusers proved were indeed sockpuppets or sockpuppeters and others of which were blocked for invility and personal attacks) and so I can understand why some would want to limit on-Wiki posting of evidence, because it's hard to resist the viciousness of some of the more disruptive editors and their socks. Anyway, I hope that helps. Best, --Le Grand Roi des CitrouillesTally-ho! 00:48, 19 November 2007 (UTC)
- It seems to me (and quite possibly many others) that the main problem here is the patent lack of transparency in regards to Durova's sockpuppet "investigative techniques". I can see Durova's logic in not revealing methodology on-wiki, but why can't she provide a syllabus of it through e-mail to established users who ask, especially users who are very logical and will all but eliminate the falsity coming from these methods? !! has many, many DYKs and should be treated like a jewel, not like a criminal in disguise. What's more unnerving is the issue Alison brought up: What if a sporadically-editing, obscure, new user was one of the "false positives"? They very well could have been the next Newyorkbrad. For that reason more than any else, Durova, in my opinion, needs to recuse herself from sockpuppet hunting until her methodology for which has been reviewed by a sensible user. —Animum (talk) 01:20, 19 November 2007 (UTC)
- Normally I do offer to circulate this type of report among trusted users upon request. Most editors who submit private evidence do not make such an offer, nor do they make it known that they have submitted any evidence privately. In this instance I stopped doing so almost immediately because new information came in that showed up a flaw in the result. DurovaCharge! 12:13, 19 November 2007 (UTC)
- Maybe instead of calling for her recusal, we could ask that she discuss with the Arbitration Committee her investigative blocks prior to blocking? --Iamunknown 01:26, 19 November 2007 (UTC)
- Iamunknown, I've already pledged to route future investigations of this type through them and let them handle it. Or, if elected, I'll let another arbitrator act upon my investigations as proof to the community that I'm not being the lone ranger. Incidentally, TOR nodes keep posting cricitisms of my actions to this thread. It's become a rather good honeypot for that purpose. ;) DurovaCharge! 01:38, 19 November 2007 (UTC)
- It appears User:!! has been ennulled unblocked[3] with an explaination. I doubt there is any one here who hasn't made well intentioned errors. Lets close this and focus efforts on more productive channels--Hu12 (talk) 01:45, 19 November 2007 (UTC)
- Agreed, and with a note to Krimpet: please do not restore posts that originate from TOR nodes.[4] I think it's a fair assumption that the person would use a legitimate account if he or she had any. As my previous post explains, the blanking was not accidental. DurovaCharge! 01:53, 19 November 2007 (UTC)
- Why remove the post, when this only gives trolls more to lap up? It would make more sense to rebut the post, or at least leave it be and let people decide for themselves if it's just baseless trolling. Though, if this user was in fact blocked by you as a sockpuppet using similar proprietary detection methods which have now proven to be faulty in at least one instance, I think it's reasonable to at least acknowledge this person's concerns. --krimpet⟲ 02:29, 19 November 2007 (UTC)
- The above post was one of two TOR node posts that Krimpet restored to this thread today. Here's the other.[5] If the user had reasonable and legitimate concerns then, after the first blanking if not before, he or she would surely have signed onto a legitimate account. Most of my investigations and sitebans have stood up to extensive scrutiny, and in unusual instances such as this I correct myself swiftly. Per WP:DUCK, TOR node posts to a thread such as this are very unlikely to be legitimate. Please do not restore such posts again without consensus support. DurovaCharge! 02:38, 19 November 2007 (UTC)
- It seems to me that there is one reason why the user in question might well need to use TOR: if their IP was blocked about a week ago. It seems to me that that is, according to the diffs actually exactly what the IPs complaining about! At least in this case, TOR node posts to a thread such as this are not unlikely to be legitimate. This is precisely the kind of slapdash thinking/repeating of conventional-wisdom-talking-points we cannot afford. Relata refero (talk) 06:10, 19 November 2007 (UTC)
- The user has been referred to the proper venues: ArbCom and the Foundation. The individual's refusal to pursue normal options does not validate the complaint in the slightest. It would be a strange day when the refusal to seek a legitimate unblock becomes an excuse to violate WP:SOCK. DurovaCharge! 17:54, 19 November 2007 (UTC)
- It seems to me that there is one reason why the user in question might well need to use TOR: if their IP was blocked about a week ago. It seems to me that that is, according to the diffs actually exactly what the IPs complaining about! At least in this case, TOR node posts to a thread such as this are not unlikely to be legitimate. This is precisely the kind of slapdash thinking/repeating of conventional-wisdom-talking-points we cannot afford. Relata refero (talk) 06:10, 19 November 2007 (UTC)
- The above post was one of two TOR node posts that Krimpet restored to this thread today. Here's the other.[5] If the user had reasonable and legitimate concerns then, after the first blanking if not before, he or she would surely have signed onto a legitimate account. Most of my investigations and sitebans have stood up to extensive scrutiny, and in unusual instances such as this I correct myself swiftly. Per WP:DUCK, TOR node posts to a thread such as this are very unlikely to be legitimate. Please do not restore such posts again without consensus support. DurovaCharge! 02:38, 19 November 2007 (UTC)
- Why remove the post, when this only gives trolls more to lap up? It would make more sense to rebut the post, or at least leave it be and let people decide for themselves if it's just baseless trolling. Though, if this user was in fact blocked by you as a sockpuppet using similar proprietary detection methods which have now proven to be faulty in at least one instance, I think it's reasonable to at least acknowledge this person's concerns. --krimpet⟲ 02:29, 19 November 2007 (UTC)
- Agreed, and with a note to Krimpet: please do not restore posts that originate from TOR nodes.[4] I think it's a fair assumption that the person would use a legitimate account if he or she had any. As my previous post explains, the blanking was not accidental. DurovaCharge! 01:53, 19 November 2007 (UTC)
- It seems to me (and quite possibly many others) that the main problem here is the patent lack of transparency in regards to Durova's sockpuppet "investigative techniques". I can see Durova's logic in not revealing methodology on-wiki, but why can't she provide a syllabus of it through e-mail to established users who ask, especially users who are very logical and will all but eliminate the falsity coming from these methods? !! has many, many DYKs and should be treated like a jewel, not like a criminal in disguise. What's more unnerving is the issue Alison brought up: What if a sporadically-editing, obscure, new user was one of the "false positives"? They very well could have been the next Newyorkbrad. For that reason more than any else, Durova, in my opinion, needs to recuse herself from sockpuppet hunting until her methodology for which has been reviewed by a sensible user. —Animum (talk) 01:20, 19 November 2007 (UTC)
(blanked 2 posts by sock of Amorrow)
- I see this IP doesn't come from a TOR node. A separate thread at AN is discussing the issue of TOR blankings. If this is a good faith post, then please overwrite the signature from your regular account or IP address. The two posts above were the second and third edits ever from that IP, and in order to reduce a drama-ridden thread it would be better to have clear accountability.[6] DurovaCharge! 03:49, 19 November 2007 (UTC)
I am still concerned about this incident. It is a not-widely-known statistical fact that if you have a population with 5% incidence of a condition, and a test that is 95% accurate in detecting the condition, a person who is selected at random and tests positive has only a 50 percent chance of actually having that condition. In other words, running a "test" on random users will result in an unacceptably high number of false positives. Durova, do you realize this? Kla’quot (talk | contribs) 03:48, 19 November 2007 (UTC)
- There's a flawed assumption in that analysis: what makes you suppose I run these tests randomly? The other due diligence methods pursued, and methods to improve them further, have already been discussed at length here. Suffice it to say that false positives are rare, I make diligent efforts to make them still rarer, and I correct errors promptly. What more can you reasonably ask? DurovaCharge! 03:52, 19 November 2007 (UTC)
- My question, which you did not answer, was a question, not an assumption. You don't have to get defensive about it, and I would still appreciate an answer. Can you clarify what you mean by "false positives are rare" - do you have a number that you can share with us?Kla’quot (talk | contribs) 04:00, 19 November 2007 (UTC)
- What I responded to was the query, In other words, running a "test" on random users will result in an unacceptably high number of false positives. Durova, do you realize this? The whole question is problematic on several levels. It simply doesn't correlate to what I actually do well enough to be answerable. Kinda like saying, "Are you aware of the risks that commuter trains pose to your daily commute?" to someone who works from home and doesn't live in a region serviced by commuter trains, and doesn't particularly want to discuss how or where they get their work done. DurovaCharge! 04:34, 19 November 2007 (UTC)
- I doubt it's possible to make a meaningful assessment of your rate of false positives, given the samples available and the potential for future cases to differ materially from past ones due to puppeteers changing their behavior. Christopher Parham (talk) 05:24, 19 November 2007 (UTC)
- (ec)(Reply to Durova) Actually it's like asking that question of someone who's arrived at work with what looks like a train ticket sticking out of their pocket. You've said, in a strangely roundabout way, that you don't pick users at random and run tests on them. That's great, however you still haven't reassured me about whether the underlying statistical phenomenon, which comes up time and time again in things like cancer screening and drug testing policy, manifests itself in the kind of evidence that you collect. The math is so counter-intuitive that the vast majority of people get it wrong. I'm not questioning your intentions or your character, which in 100% of my previous interactions with you has never seemed less than stellar. Designing test methods in a way that will give a low and quantifiable number of false positives is difficult. Knowing that someone as smart as you misread the evidence, I'm afraid that giving the same evidence to ArbCom is likely to have the same result. Kla’quot (talk | contribs) 05:41, 19 November 2007 (UTC)
- Well, without getting into too many specifics I'll lay out some differences between this and my arbcom evidence. For the Alkivar case I was preparing a three part presentation. The first was on Burntsauce, the second on Alkivar, and the third didn't fully materialize because the research generated some unintended results. During the third part of the investigation I implemented what appeared to be a routine block that precipitated the the Eyrian arbitration case. Eyrian's own conduct in the aftermath overshadows the work I was doing in that portion (except for the Dannycali block). The Alkivar portion of that presentation was a six page text file of description and dozens of diffs, distilled from original notes that ran to about 30 pages, and the Alkivar evidence itself implemented some more definitive methodologies. The Burntsauce evidence was on a par with that. So while I wouldn't call two pages and 28 diffs superficial or hasty, it wasn't as extensive or as conclusive. With regard to statistical analysis, I don't think there are enough stable variables here or a large enough sample size to really make that useful. What matters from my perspective is that this does generate mostly accurate results and we've been good about correcting the errors swiftly. I'd love to get the false positives down to zero. Everyone would. It takes a lot of hard work to locate and implement specific improvements. DurovaCharge! 06:04, 19 November 2007 (UTC)
- I don't know why you're bringing up the previous Arbcom case. I didn't ask about that case and have never gone anywhere near it. I am disappointed that your response when asked to clarify what you meant by "rare" was to repeat your assertion with its original level of vagueness. I only hope that the people whom you ask to act on your evidence will read my concerns and understand them more than you do. Kla’quot (talk | contribs) 09:01, 19 November 2007 (UTC)
- I apologize if that response missed the point of your question. We seem to be operating from different paradigms. If I understand you, you want a statistical clarification. I've attempted to convey that the question is framed in ways that make not useful. If that's an unsatisfactory answer then the basic dilemma is this: in order to begin answering it I'd have to correlate it more closely to the actual field conditions it attempts to describe. You seem to regard that as digressive and I definitely regard it as treading on sensitive territory. Then if we got that far, we'd be stuck with too small a sample size to parse statistically. Comparing two different sock investigations can be like comparing apples and oranges. What I want to ask in return is why you place such weight on framing the question in these terms when the good results are valuable, the false positives are uncommon, and the errors get corrected quickly. DurovaCharge! 09:58, 19 November 2007 (UTC)
- I do appreciate your response. Let me try this again: I asked two compltely separate questions. : First question: "It is a not-widely-known statistical fact that if you have a population with 5% incidence of a condition, and a test that is 95% accurate in detecting the condition, a person who is selected at random and tests positive has only a 50 percent chance of actually having that condition." This is a mathematical fact. I asked you whether you are aware that this fact exists. I was expecting something like, "I completely understand what you are talking about" or "I didn't know that." Second question: You have repeatedly asserted that false positives are "rare." What is your definition of rare? I was hoping for an answer like, "In the past 6 months my methods have led to blocks of 100 accounts and only 1 of them has asked to be unblocked." Please do not reply by saying that false positives are rare. Kla’quot (talk | contribs) 10:31, 19 November 2007 (UTC)
- You appear to be asking that Durova be perfect. Why? One mistake is now known, and it was swiftly rectified. 99 blocks which were not mistakes, stand. Other blocks in which Durova's investigation played a part, also stand. The group of individuals most focused on abusing Wikipedia right now are resourceful and determined. It will not be especially helpful to tell them the ways in which they betray themselves, for obvious reasons. Of course the best solution for all concerned would be for them to go away and leave us alone, but since they refuse to do that, and since they are resourceful, determined, devious and obsessive there are likely to be very occasional errors. As long as they are repaired swiftly and we both apologise and learn from them I fail to see the need for further debate. Sure, it would be better if such false positives never happened. The way to get there is by the banned abusers going away. Since we can't make them do that, all we can do is watch. You saw the damage Burntsauce did, and that also had the regrettable side-effect of bringing down Alkivar. Nobody is happy about that. For reasons I can't fathom, a few people seem to give encouragement to these banned users off-wiki. For reasons I also can't fathom, but which may not be entirely unrelated, we seem to be engaging in a witch-hunt here. Apart from apologising and swiftly reversing the block, what else is Durova supposed to do? Guy (Help!) 11:41, 19 November 2007 (UTC)
- You seem to not acknowledge that the community has almost universally expressed concern that this particular block was of an editor with several positive contributions, indeed someone who appears a near-ideal user. If it happened to someone with a more dicey record, like 90% of the users on the 'pedia, most of us would have let it slide. What this means is that of the 99 blocks that 'stand', in your reasoning, we have no way of knowing for certain, absent the normal discussion and scrutiny, that they are in fact all justified. They may well be, and (in my opinion) probably are; but we can't assume that. We have checks and balances, scrutiny and noticeboards, for precisely this reason. If we start seeing 'banned users' or their enablers under every bush in a paranoid manner, the project will suffer. And it appears that that's what happening. This is not the first time that this is being said. If you are genuinely interested in reducing drama, then you need to take the community along with you. It seems you haven't, or have gone too far. Can we agree on that? Relata refero (talk) 12:33, 19 November 2007 (UTC)
- Why would I need to "acknowledge" something that Durova made clear right up front, with the original post to this noticeboard? The correct response here was what Durova did, which was to reverse the block, apologise, and learn from the experience. It is not clear to me what else is expected from this continued debate. The cause of the problem is very clear: long-term abuse by determined and unscrupulous people. Be on your guard, everyone, and "trust but verify". I really don't see what this debate is achieving at this point. Guy (Help!) 13:38, 19 November 2007 (UTC)
- No, dash it, Guy, what you need to acknowledge is our concern that the only reason that the unblock happened this time is because the user's as near ideal as can be. All of us are worried that if it wasn't an ideal user, we wouldn't even be here. (We're also a little guilty, perhaps.) The cause of this problem is not long-term abuse, it is overzealous enforcement. We all know that one extreme is as bad as the other, history and experience will tell all of us so, and the community seems to think we've swung too far towards the enforcement extreme. So lessen up a bit, OK? This is more drama than any of us want. Also we don't want too many productive users scared off the project. Relata refero (talk) 14:08, 19 November 2007 (UTC)
- Per this check, I average about one apology a month for all purposes, whether block related or not. That's roughly one every 1000 edits. And with regard to the first question, I'm stunned to see a pop quiz in basic statistics here that wasn't intended analogously. DurovaCharge! 12:03, 19 November 2007 (UTC)
- You appear to be asking that Durova be perfect. Why? One mistake is now known, and it was swiftly rectified. 99 blocks which were not mistakes, stand. Other blocks in which Durova's investigation played a part, also stand. The group of individuals most focused on abusing Wikipedia right now are resourceful and determined. It will not be especially helpful to tell them the ways in which they betray themselves, for obvious reasons. Of course the best solution for all concerned would be for them to go away and leave us alone, but since they refuse to do that, and since they are resourceful, determined, devious and obsessive there are likely to be very occasional errors. As long as they are repaired swiftly and we both apologise and learn from them I fail to see the need for further debate. Sure, it would be better if such false positives never happened. The way to get there is by the banned abusers going away. Since we can't make them do that, all we can do is watch. You saw the damage Burntsauce did, and that also had the regrettable side-effect of bringing down Alkivar. Nobody is happy about that. For reasons I can't fathom, a few people seem to give encouragement to these banned users off-wiki. For reasons I also can't fathom, but which may not be entirely unrelated, we seem to be engaging in a witch-hunt here. Apart from apologising and swiftly reversing the block, what else is Durova supposed to do? Guy (Help!) 11:41, 19 November 2007 (UTC)
- I do appreciate your response. Let me try this again: I asked two compltely separate questions. : First question: "It is a not-widely-known statistical fact that if you have a population with 5% incidence of a condition, and a test that is 95% accurate in detecting the condition, a person who is selected at random and tests positive has only a 50 percent chance of actually having that condition." This is a mathematical fact. I asked you whether you are aware that this fact exists. I was expecting something like, "I completely understand what you are talking about" or "I didn't know that." Second question: You have repeatedly asserted that false positives are "rare." What is your definition of rare? I was hoping for an answer like, "In the past 6 months my methods have led to blocks of 100 accounts and only 1 of them has asked to be unblocked." Please do not reply by saying that false positives are rare. Kla’quot (talk | contribs) 10:31, 19 November 2007 (UTC)
- I apologize if that response missed the point of your question. We seem to be operating from different paradigms. If I understand you, you want a statistical clarification. I've attempted to convey that the question is framed in ways that make not useful. If that's an unsatisfactory answer then the basic dilemma is this: in order to begin answering it I'd have to correlate it more closely to the actual field conditions it attempts to describe. You seem to regard that as digressive and I definitely regard it as treading on sensitive territory. Then if we got that far, we'd be stuck with too small a sample size to parse statistically. Comparing two different sock investigations can be like comparing apples and oranges. What I want to ask in return is why you place such weight on framing the question in these terms when the good results are valuable, the false positives are uncommon, and the errors get corrected quickly. DurovaCharge! 09:58, 19 November 2007 (UTC)
- I don't know why you're bringing up the previous Arbcom case. I didn't ask about that case and have never gone anywhere near it. I am disappointed that your response when asked to clarify what you meant by "rare" was to repeat your assertion with its original level of vagueness. I only hope that the people whom you ask to act on your evidence will read my concerns and understand them more than you do. Kla’quot (talk | contribs) 09:01, 19 November 2007 (UTC)
- Well, without getting into too many specifics I'll lay out some differences between this and my arbcom evidence. For the Alkivar case I was preparing a three part presentation. The first was on Burntsauce, the second on Alkivar, and the third didn't fully materialize because the research generated some unintended results. During the third part of the investigation I implemented what appeared to be a routine block that precipitated the the Eyrian arbitration case. Eyrian's own conduct in the aftermath overshadows the work I was doing in that portion (except for the Dannycali block). The Alkivar portion of that presentation was a six page text file of description and dozens of diffs, distilled from original notes that ran to about 30 pages, and the Alkivar evidence itself implemented some more definitive methodologies. The Burntsauce evidence was on a par with that. So while I wouldn't call two pages and 28 diffs superficial or hasty, it wasn't as extensive or as conclusive. With regard to statistical analysis, I don't think there are enough stable variables here or a large enough sample size to really make that useful. What matters from my perspective is that this does generate mostly accurate results and we've been good about correcting the errors swiftly. I'd love to get the false positives down to zero. Everyone would. It takes a lot of hard work to locate and implement specific improvements. DurovaCharge! 06:04, 19 November 2007 (UTC)
- (ec)(Reply to Durova) Actually it's like asking that question of someone who's arrived at work with what looks like a train ticket sticking out of their pocket. You've said, in a strangely roundabout way, that you don't pick users at random and run tests on them. That's great, however you still haven't reassured me about whether the underlying statistical phenomenon, which comes up time and time again in things like cancer screening and drug testing policy, manifests itself in the kind of evidence that you collect. The math is so counter-intuitive that the vast majority of people get it wrong. I'm not questioning your intentions or your character, which in 100% of my previous interactions with you has never seemed less than stellar. Designing test methods in a way that will give a low and quantifiable number of false positives is difficult. Knowing that someone as smart as you misread the evidence, I'm afraid that giving the same evidence to ArbCom is likely to have the same result. Kla’quot (talk | contribs) 05:41, 19 November 2007 (UTC)
- My question, which you did not answer, was a question, not an assumption. You don't have to get defensive about it, and I would still appreciate an answer. Can you clarify what you mean by "false positives are rare" - do you have a number that you can share with us?Kla’quot (talk | contribs) 04:00, 19 November 2007 (UTC)
Durova, to refuse to answer some of the editor's questions here and say that if anyone has a problem with your methods to "take it to the ArbCom" isn't very helpful. The overworked ArbCom doesn't have to get involved if you'll answer to the concerned editors on this page. Would you please answer the following questions?
- How do you decide who will be a target of your sleuthing?
- What methods do you use to sleuth editors?
- What is your rate of false positives and how do you know when you get a false positive?
- Who are the ArbCom members that you mail your evidence to, or is it to the entire ArbCom?
- Why do you feel that you should act "behind the scenes" with the ArbCom over these matters, when you could just as well post your observations and proposed actions here or at AN, which I've seen other admins do?
- Do you share your evidence with the accused editor and ask for an explanation before taking action on their account? Cla68 (talk) 08:15, 19 November 2007 (UTC)
- In light of these previous questions there's some room for doubt about whether that list is assembled in good faith. If it was in good faith, could you clarify? Why do you repeat a series of questions, most or all of which I've provided reasonable justification for declining to answer onsite, without any new reason for answering them? DurovaCharge! 09:58, 19 November 2007 (UTC)
Your "justification" may seem reasonable to yourself: not to many others. You appear to have appointed yourself as the Wikipedia Bureau of Investigation, and treat the criticisms of little admins with disdain. As I have mentioned to you on several occasions, if you methods are so secret, for @@@@'s sake don't shout about them! Otherwise we might just conclude that your spouting self-agrandiziung rubbish which is detrimental to the project. Given your repeted blanking of contributions to this page (four since 10:00 UTC today), I have warned you to stop. If you continue to user your editing privileges in such a controversial manner, you will only have yourself to blamew if they are suspended. Physchim62 (talk) 12:34, 19 November 2007 (UTC)
- If any one doubts that Durova is in the business of self-agrandisement, they should take a look at this edit from her talk page, coming just two minutes after she semiprotected it (for the first time). Physchim62 (talk) 13:56, 19 November 2007 (UTC)
This discussion is beginning to generate more heat than light. shoy (words words) 14:28, 19 November 2007 (UTC)
- Fair enough, we can always move it to Wikipedia:Requests for comment/Durova if you think that would be more helpful. Physchim62 (talk) 18:24, 19 November 2007 (UTC)